Call for participation: task force to work on resource to allow users to quickly decide on what simulator/tool they want to use

[sanjayankur31]

Please see this ticket for more information on the task:

https://github.com/OCNS/SoftwareWG/issues/61

We would like to set up a task force to deliver the first version of a resources that allows users to quickly decide on what simulator/toll they should/want to use. Please comment on this ticket if you would like to join the task force.

[sanjayankur31]

I can't remember if we've run into "biosimulators" yet? Also has a simulation selection page:

https://biosimulators.org/simulators

[brenthuisman]

@sanjayankur31 They (never heard of this initiative) have an interesting checklist for database entries (hope the link works): click

They seem to define a set of interface conventions, e.g. simulator interfaces and there's a whole bunch of metadata they collect/use (such as use of various ontology terms). Pity that they don't seem to be aware of the one that's used in Ebrains.

Curious how some simulators are in there, where they don't have the required metadata in repo. Maybe @mstimberg knows how they added Brian?

[sanjayankur31]

Yeh, this is from the systems biology side of things related to the COMBINE initiative. I think the list was generated some time ago when the paper was in the works and Brian could be there through their NeuroML connection (NeuroML being a COMBINE standard), but I'm not sure.

[brenthuisman]

I haven't searched that long, but I can't find where they store that meta-data. Before I prep a whole bunch of meta-data for Arbor, it'd be good to know if they include things transitively (Arbor also can execute NeuroML simulations).

If we can find that meta-data (through their interface I can't see e.g. output formats), it could be a good base to start actually. Maybe we could write our tool around their meta-data?

Also what they list as dependencies for NEURON appear to be dependants.

[mstimberg]

Hi everyone, I did not know of biosimulators before and wasn't involved in getting Brian in there – but I also suspect it has something to do with COMBINE/NeuroML. They put in quite a bit of work in the metadata and all the ontology stuff, although I am not quite sure how useful these lists like https://www.ebi.ac.uk/ols/ontologies/kisao and https://www.ebi.ac.uk/ols/ontologies/sbo actually are in practice. I guess I see the value in having a precise ontology term for, say, the "forward Euler method", but while e.g. a single-compartment HH-like model is technically a non-spatial continuous framework, this doesn't quite sound like the way you'd describe it in computational neuroscience. More importantly, mechanisms like LEMS, NMODL, Brian's equation language, etc. let you build almost arbitrary things with a simulator, so I don't know how you'd classify that.

Regarding the metadata and using it for our own purposes: this seems to be tricky, since what they store is not (only) the metadata for the tool itself but for their (CLI and Python) interface wrapper around it. E.g. for Brian, it lists SED-ML as simulation formats and COMBINE OMEX as archive formats, but Brian itself doesn't support these. It also lists pyLEMS/pyNeuroML as dependencies, but they are of course only dependencies when you want to run NeuroML/LEMS models. In general, if you actually want to run your models in the way that biosimulators supports (executing a SED-ML simulation description that refers to NeuroML/LEMS files), then it doesn't make much sense to use Brian for that (except maybe for cross-checking results from other simulators) – you'll lose most of Brian's features that make it interesting.

[mstimberg]

Hi everyone, just to make sure we are all on the same page: We are having our next meeting in the coming week, on Wednesday Feb 8th, at 1400 UTC (time in your timezone).

:tv: link: https://cnrs.zoom.us/j/96162620842?pwd=eEpVQ1gwL3p0R2UvRk9GdEJ6eG9SZz09

As a reminder: our "homework" is to

1. Come up with some sort of sketch/prototype of a proposed interface (can be a simple hand-made drawing or a fully working dashboard with fake data – you decide :laughing: )
2. think about the technology, what kind of framework could/should we use?

Notes from our last meeting are here: https://hackmd.io/@mstimberg/rkN-yS6ji "Master" notes document is here: https://hackmd.io/@mstimberg/HJXqYPMPo

See you next week :wave:

[mstimberg]

Quick reminder that we'll have our upcoming meeting in ~40 mins (meeting time in your timezone). :tv: https://cnrs.zoom.us/j/96162620842?pwd=eEpVQ1gwL3p0R2UvRk9GdEJ6eG9SZz09 :notebook: https://hackmd.io/@mstimberg/S1PxWQZ6j

[mstimberg]

We started the meeting, please join if you can (https://cnrs.zoom.us/j/96162620842?pwd=eEpVQ1gwL3p0R2UvRk9GdEJ6eG9SZz09).

[sanjayankur31]

Repo created here: https://github.com/OCNS/simselect

I've invited all you folks to the team which has maintenance access to the repo. Please do let me know if you haven't received your invitation.

[brenthuisman]

Hi all, sorry for missing the sesh yesterday, down and out unfortunately. Will be there next time.

[sanjayankur31]

I folks, sorry, i completely didn't realise. I'm at NeuroDataShare most of next week, so I'll probably miss our meeting. I'll make sure my tasks from last meeting are done and I'll look at the logs and catch up :+1:

[mstimberg]

:alarm_clock: Reminder that we'll have our next task force meeting later today, at 1400 UTC (time in your timezone) Please have a look at last week's meeting notes to see what we decided last time.

:tv: https://cnrs.zoom.us/j/96314388091?pwd=T3F1d2lSeWdPWGlHODhkclhuQkhLZz09

[mstimberg]

Hi again, the meeting is about to start, meeting notes will be here: :notebook: https://hackmd.io/@mstimberg/r1w_pqXAs

[mstimberg]

:alarm_clock: Reminder that we'll have our next task force meeting later today, at 1400 UTC (time in your timezone)

I updated the last week's meeting notes with our discussion, please have a look and of course feel free to add/edit if I forgot something.

Regarding the feasibility of exporting a panel website statically with their save/embed functionality: I looked into this a bit, and unfortunately it seems that this functionality is limited to a number of classes such as DiscreteSlider and Select that allow for the selection of single values, but the more complex widgets such as MultiChoice or IntRangeSlider are not supported (their _supports_embed attribute is False). I guess this is because their state space quickly becomes large. Let's discuss during the meeting what that means for the way forward.

Please note that @ree-gupta converted our current (still very rough) tools list to yaml :pray: Very useful to play around with: https://github.com/OCNS/simselect/blob/main/simtools/simtools.yaml

See you in a bit, notebook and zoom link below:

:notebook: https://hackmd.io/@mstimberg/SkaUQ181n :tv: https://cnrs.zoom.us/j/99113116277?pwd=bDhSaDdZQ1lCazEvcGdJcWdJY1podz09

[brenthuisman]

Thanks @ree-gupta and @mstimberg! I have to skip today's meet though, see you in two weeks!

[mstimberg]

Thanks @ree-gupta and @mstimberg! I have to skip today's meet though, see you in two weeks!

Sure, no worries. I'll update the notes so you can read up later. And of course we'll volunteer you for all the boring tasks in your absence :wink:

[mstimberg]

I just started the meeting, please join us if you can :blush: https://cnrs.zoom.us/j/99113116277?pwd=bDhSaDdZQ1lCazEvcGdJcWdJY1podz09

[mstimberg]

:alarm_clock: Reminder that we'll have our next task force meeting this Wednesday, at 1400 UTC (time in your timezone)

Please have a look at last week's meeting notes to remind you of what we discussed.

@sanjayankur31 actually, nevermind the docker file – for Python projects, fly can automatically generate a container based on the requirements.txt file. With the CLI, it was trivial to deploy (but we need to put this into a GitHub Action): https://simselect.fly.dev/ :rocket:

[sanjayankur31]

Ah, that's great. See you all at the meeting!

[mstimberg]

Sorry, I'm a bit late :grimacing: Zoom link here: :tv: https://cnrs.zoom.us/j/92435856592?pwd=R1o4dEcxTytIOHI0V2pJZ2g0N3d0UT09

[mstimberg]

Our meeting notes will be here: :notebook: https://hackmd.io/jGUGUDCrT8abdGtJOAKVPw

[dissagaliyeva]

Ah, sorry guys, can't make it! Having a work call :(

[mstimberg]

:alarm_clock: Reminder that we are having our next task force meeting in ~2h, i.e. today at 1400 UTC (time in your timezone). Please be aware that in many places local time switched to day-time saving times :sun_with_face:, so this meeting might be one hour later than usual (it is for me!).

You can find last week's meeting notes here: https://hackmd.io/jGUGUDCrT8abdGtJOAKVPw

For today's meeting: :notebook: Meeting notes: https://hackmd.io/@mstimberg/Bkvkvyjb2 :tv: Zoom link: https://cnrs.zoom.us/j/96879633591?pwd=aWxXcGFlVG01YWRDbnQ3cmpjUkJ2QT09

[elianecr]

Today I won't be able to participate, I'm in the final adjustments of my PhD thesis, running out of time!!! I'm sorry...

[mstimberg]

Today I won't be able to participate, I'm in the final adjustments of my PhD thesis, running out of time!!! I'm sorry...

Sure thing, no worries.

We had a short meeting (see notes at https://hackmd.io/@mstimberg/Bkvkvyjb2), and decided (if that works for everyone of course) to go back to the old meeting time by shifting it to 1300 UTC (next meeting April 19th).

[mstimberg]

:alarm_clock: Reminder that we are having our next task force meeting tomorrow, i.e. Wednesday Apr 19th at 1300 UTC (time in your timezone).

Last week's meeting notes are here: https://hackmd.io/@mstimberg/Bkvkvyjb2

[mstimberg]

Another reminder, we'll have our meeting today (at 1300 UTC, i.e. in roughly 3.5 hours: time in your timezone).

I now set this up as a recurring meeting (:calendar: import it into your calendar), so the Zoom link should stay the same from now on: :tv: https://cnrs.zoom.us/j/94102303309?pwd=SkIzOUJSVG1aWURROE5rR3d6Yi9iQT09

Here's the link to the notes, please feel free to add agenda items :notebook: https://hackmd.io/@mstimberg/SkqazEpz3

Looking forward to seeing many of you there!

[mstimberg]

I started the meeting, please join if you can! :tv: https://cnrs.zoom.us/j/94102303309?pwd=SkIzOUJSVG1aWURROE5rR3d6Yi9iQT09

[brenthuisman]

Whoops, forgot to join yesterday. Thanks for the calendar file @mstimberg ; added!

[mstimberg]

:wave: hi everyone, reminder that we are having another meeting later today, at 1300 UTC (time in your timezone). The Zoom link is the same as last time: :tv: https://cnrs.zoom.us/j/94102303309?pwd=SkIzOUJSVG1aWURROE5rR3d6Yi9iQT09 Here are the notes from last week, please feel free to add agenda items to today's meeting notes: :notebook: https://hackmd.io/@mstimberg/B1R6f2JN3

Hope to see many of you there!

[brenthuisman]

Hi all, I'm going to be late, but I'll try to catch the end.

[mstimberg]

Hi all, I started the meeting, please join me if you can: :tv: https://cnrs.zoom.us/j/94102303309?pwd=SkIzOUJSVG1aWURROE5rR3d6Yi9iQT09

[sanjayankur31]

@mstimberg @brenthuisman @ree-gupta @elianecr @dissagaliyeva (and everyone else): sorry, I haven't really had a chance to work on simselect recently. Are we still on for a meeting today to maybe distribute tasks and re-gain some momentum? :)

[mstimberg]

:wave: Hi everyone. It's been a while, but please be reminded that we are having another meeting today, at 1300 UTC (time in your timezone). The Zoom link is the same as last time: :tv: https://cnrs.zoom.us/j/94102303309?pwd=SkIzOUJSVG1aWURROE5rR3d6Yi9iQT09 Meeting notes will be here: :notebook: https://hackmd.io/@mstimberg/HJ5_en4Ih

Hope to see many of you there!

[mstimberg]

Are we still on for a meeting today to maybe distribute tasks and re-gain some momentum? :)

Cross-over post :laughing: Yes, let's have another meeting later today!

[sanjayankur31]

haha, awesome, see you all in a bit.

[mstimberg]

I started the Zoom, please join if you can :blush: :tv: https://cnrs.zoom.us/j/94102303309?pwd=SkIzOUJSVG1aWURROE5rR3d6Yi9iQT09

[ree-gupta]

Hey everyone! I apologize for not being able to join yesterday and for my lack of progress on simselect lately. I'm sorry again! Rest assured, I will try to make some progress on the assigned issues, and meet you all on the 14th

PS - Allen institute is conducting a community survey on neural modelling software - https://www.surveymonkey.com/r/2023NeuralModelingSoftware We could keep an eye on the results to later improvise simselect

[sanjayankur31]

PS - Allen institute is conducting a community survey on neural modelling software - https://www.surveymonkey.com/r/2023NeuralModelingSoftware We could keep an eye on the results to later improvise simselect

I filled that out yesterday. It's quite specific to the Allen tools that the grant covers---whether people are using BTMK/VND and so on.

We (the Software WG) had done a more general survey before to get some idea of what tools people use. Maybe we could look at this again just as a reminder. Results etc here:

https://github.com/OCNS/SoftwareWG/issues/82

[ree-gupta]

Hi Ankur! Thank you. This looks cool. I was not aware that the software WG already conducted a survey. Insightful!

[sanjayankur31]

Folks, unfortunately, I have a clash tomorrow---we have a departmental seminar at the same time---so I won't make it to our meeting. I made a few tweaks to the repo and opened a PR adding descriptions to all the simtools files. Will work on other bits as I find more time too.

[mstimberg]

:wave: Hi everyone. Sorry to hear that you can't join @sanjayankur31 (but thanks for your changes!). For everyone else, here's the usual reminder that we are having another meeting tomorrow, at 1300 UTC (time in your timezone). The Zoom link is the same as last time: tv https://cnrs.zoom.us/j/94102303309?pwd=SkIzOUJSVG1aWURROE5rR3d6Yi9iQT09

The website did evolve a bit since we talked last time, please take a look at the current state: https://simselect.fly.dev/ By accident, I made a commit directly to main that I wanted to discuss first – I replaced the modal overlay window for the details by a simple area below the list of simulators. The advantage of my mistake is that you can easily have a look at the website to see what it looks like ;-) Let me know what you think in the respective issue (OCNS/simselect#26) or tomorrow in the meeting.

[brenthuisman]

Hi all, I also have to give my apologies for today. But I really like how the website is shaping up!

[mstimberg]

Hi, we are in the meeting, if you can please join!

:tv: Zoom: https://cnrs.zoom.us/j/94102303309?pwd=SkIzOUJSVG1aWURROE5rR3d6Yi9iQT09 :notebook: Meeting notes: https://hackmd.io/@mstimberg/HJ_ht4PDn

[mstimberg]

The usual reminder: we are having our taskforce meeting today, at 1300 UTC (time in your timezone).

The zoom link is unchanged: :tv: https://cnrs.zoom.us/j/94102303309?pwd=SkIzOUJSVG1aWURROE5rR3d6Yi9iQT09

:notebook: Meeting notes: https://hackmd.io/@mstimberg/SJI_FiY_2

[mstimberg]

:point_up: Sorry about not sending this earlier, I forgot to click on the button :roll_eyes:

We are in the zoom now, please join if you can.

[ree-gupta]

Hey all! I will unfortunately not be able to attend the meeting today, but will definitely go over the meeting notes afterwards. Hope to see some of you in Leipzig in a few days. :)

[mstimberg]

Late, but hopefully not too late, here's the usual reminder: we are having our taskforce meeting today, at 1300 UTC (time in your timezone).

The zoom link is unchanged: :tv: https://cnrs.zoom.us/j/94102303309?pwd=SkIzOUJSVG1aWURROE5rR3d6Yi9iQT09

:notebook: Meeting notes: https://hackmd.io/@mstimberg/BJMUZX3Fh

Hey all! I will unfortunately not be able to attend the meeting today, but will definitely go over the meeting notes afterwards. Hope to see some of you in Leipzig in a few days. :)

No worries, see you in a few days!

[sanjayankur31]

We released v0.1.0 at the meeting. Thanks everyone who has contributed :clap: :clap:

https://github.com/OCNS/simselect/releases/tag/v0.1.0

[sanjayankur31]

Just got my recurrent reminder, so thought I'd check---do we have a meeting today folks?